This paper examines the evolving landscape of AI-assisted programming and proposes a novel approach to enhance the collaboration between generative AI and human programmers. Traditional methods of code generation often lead to code blending, making it challenging to distinguish between AI-generated and human-written code. To address this issue, we introduce the concept of an AI-Assisted programming paradigm, which involves extending programming languages with specialized annotations or syntax to control AI code generation. This framework aims to separate AI-generated code from human-written code, improving code readability and maintainability.